RESEARCH

Comparison of simulated periodontal bone defect depth

measured in digital radiographs in dedicated and non-dedicated

software systems

G Scaf*,1, CE Sakakura1, PFD Kalil1, JAN Dearo de Morais1, LCM Loffredo2 and A Wenzel3

1Department of Oral Diagnosis and Surgery, Araraquara Dental School, State of São Paulo University, Unesp, Araraquara, São Paulo,
Brazil; 2Department of Social Dentistry, Araraquara Dental School, State of São Paulo University, Unesp, Araraquara, São Paulo,
Brazil; 3Department of Oral Radiology, School of Dentistry, Faculty of Health Sciences, University of Aarhus, Denmark

Objectives: To compare simulated periodontal bone defect depth measured in digital radiographs
with dedicated and non-dedicated software systems and to compare the depth measurements from
each program with the measurements in dry mandibles.
Methods: Forty periodontal bone defects were created at the proximal area of the first premolar in
dry pig mandibles. Measurements of the defects were performed with a periodontal probe in the dry
mandible. Periapical digital radiographs of the defects were recorded using the Schick sensor in a
standardized exposure setting. All images were read using a Schick dedicated software system (CDR
DICOM for Windows v.3.5), and three commonly available non-dedicated software systems (Vix
Win 2000 v.1.2; Adobe Photoshop 7.0 and Image Tool 3.0). The defects were measured three times
in each image and a consensus was reached among three examiners using the four software systems.
The difference between the radiographic measurements was analysed using analysis of variance
(ANOVA) and by comparing the measurements from each software system with the dry mandibles
measurements using Student’s t-test.
Results: The mean values of the bone defects measured in the radiographs were 5.07 mm,
5.06 mm, 5.01 mm and 5.11 mm for CDR Digital Image and Communication in Medicine (DICOM)
for Windows, Vix Win, Adobe Photoshop, and Image Tool, respectively, and 6.67 mm for the dry
mandible. The means of the measurements performed in the four software systems were not
significantly different, ANOVA (P ¼ 0.958). A significant underestimation of defect depth was
obtained when we compared the mean depths from each software system with the dry mandible
measurements (t-test; P ø 0.000).
Conclusions: The periodontal bone defect measurements in dedicated and in three non-dedicated
software systems were not significantly different, but they all underestimated the measurements
when compared with the measurements obtained in the dry mandibles.
Dentomaxillofacial Radiology (2006) 35, 422–425. doi: 10.1259/dmfr/61300663

Keywords: digital radiography; alveolar bone loss, measurement

Introduction

The use of digital radiography is increasing and several
systems based on either solid state sensors or photostimul-
able phosphor plates are available for dentistry with a
substantial scientific validation.1–4 Despite the advantages
offered by these technologies, such as the elimination of
chemical processing, lower radiation doses to the patient,

image processing and improved image management, the
possibility of easy and fast transmission of image data to
third parties over the internet could be considered one of
the main advantages, offering easy access for the dentist to
second opinions on diagnosis and treatment planning.5

The availability of several digital dental radiography
systems (www.odont.au.dk/rad/Digitalx.htm) has intro-
duced new problems concerning image exchange since
there is no absolute standard in dentistry for image file
format. This means that digital images acquired in one

*Correspondence to: Dr. Gulnara Scaf, Rua Humaitá, 1680, 14801-903, Araraquara,

São Paulo, Brazil; E-mail: scaf@foar.unesp.br

Received 8 August 2005; revised 19 December 2005; accepted 28 December 2005

Dentomaxillofacial Radiology (2006) 35,Dentomaxillofacial Radiology (2006) 35, 422–425
q 2006 The British Institute of Radiology

http://dmfr.birjournals.org



system are not easily imported into another system’s
software. Efforts are ongoing to implement the DICOM
(Digital Image and Communication in Medicine) standard
in dentistry to encourage interchangeability among soft-
ware systems.6 However, few dental digital systems have
offered a DICOM solution,7 which moreover may not be
the right solution for dentistry.8 Furthermore, only a minor
percentage of dental clinics have yet switched from
traditional film-based radiography to digital imaging.9,10

However, increasing numbers of dentists have a personal
computer and internet facilities in their offices and there are
several low cost software packages available on the market
to import and display digital images. These include
freeware, e.g. Image Tool 3.0 available at http://ddsdx.
uthscsa.edu/dig/itdesc.html. It may be reasonable to believe
that dedicated software that comes with a digital system,
responsible for capturing digital radiographic images,
would be optimized for image interpretation acquired and
displayed with this receptor. However, there is a lack of
studies addressing the interoperability among different
imaging software systems.

The aims of this study were to compare simulated
periodontal bone defect depth measured in digital radio-
graphs in dedicated and non-dedicated software systems
and to compare depth measurements in each program with
those in dry mandibles.

Materials and methods

Forty dry pig hemi-mandibles were used in this study.
Periodontal two-wall bone defects were created at the
mesial side of the first premolar in the apical direction
using a slow-speed round bur no. 4. The mesial side was
chosen because it is the proximal site that gives the best
access to create the periodontal bone defect since it
presents a diastema between the two contiguous teeth. The
bone defect width was determined by the bur diameter and
the defect depth was random. Bone was removed
corresponding to the bur diameter, and a small chisel was
used to refine the defect eliminating all bone chips attached
to the root surface. The buccal wall and part of the
proximal wall were removed, thus only the lingual wall
remained. A tiny mark was made at the cementoenamel

junction with a double-face diamond disc to establish the
upper border of the bone defect, and it was used as a
reference point because the detection of the radiographic
image of the cementoenamel junction was difficult.

Image acquisition
Digital images of the defects were acquired using the
CMOS Schickw sensor (Schick Technologies Inc., Long
Island, NY). The geometric alignment between the sensor
and the hemi-mandible was standardized using a fixing
device. A 9.22 mm long aluminium step wedge was fixed to
the sensor, and was also used to calculate the magnification
and distortion in the further analysis using an object with a
known length. The radiographs were taken with the vertical
long axis of the hemi-mandible fixed perpendicular to the
central ray and parallel to the sensor at a 70 cm focal spot to
object distance. The X-ray unit operated at 70 kVp, 10 mA
and 18 impulses. A wooden block of 2 cm thickness was
placed in front of the mandible to simulate soft tissue.

The images were stored in TIFF (Tagged Image File
Format) without compression (8 bits with a resolution of
600 dpi, a file of about 700 KB).

Bone defect measurement
The bone defect depth was measured as the distance from
the deepest part of the defect to the cementoenamel
junction. A periodontal probe (Williams Probe; Hu-Friedy,
Chicago, IL) was used to measure the defects in the dry pig
mandibles, positioned by hand at the bottom of the defect
with the vertical long axis parallel to the root surface. The
cementoenamel junction point could be demarcated on the
probe with the aid of a small endodontic rubber ring, and
the length between the probe tip and the rubber ring was
measured by a digital calliper (Ultra-Call Mark III; Fowler,
Switzerland) (Figure 1a). This measurement was con-
sidered the true depth (gold standard).

The digital images were evaluated in four image
software systems (Figure 1b); one, the dedicated Schick
program CDR DICOM for Windows v.3.5, and another
three non-dedicated software systems: Vix Win 2000 v.1.2,
Image Tool 3.00 and Adobe Photoshop 7.0 (Table 1). The
software was installed in a laptop computer (Satellite P25-
S507; Toshiba, Taiwan), and the images were evaluated on
a 1700 WXGA (1440 £ 900) monitor.

Figure 1 Measurement in a dry pig hemi-mandible (a) using periodontal probe and endodontic rubber ring and (b) digital radiographic image showing
the distance from the deepest part of the defect to the cementoenamel junction

Bone defects in software systemsBone defects in software systems
G Scaf et al 423

Dentomaxillofacial Radiology



Prior to the measurements of the defect depths in the
digital images, all the images were standardized according
to a known length from the step wedge and the ruler tool
available in the software system.

The depth measurements were performed by three
examiners: one radiologist, one PhD student and one
undergraduate student, for both the dry mandibles and the
digital images. The examiners were calibrated before
reading sessions and performed the measurements together,
in three consecutive sessions, in dimmed room light.

Statistical analysis
The three measurements of each defect were averaged, and
a mean value was obtained. Analysis of variance
(ANOVA) was used to compare the depth measurements
from the four software systems. Student’s t-test was used to
compare the depth measurements between each method
and the dry mandible. The 95% confidence intervals for the
mean depth measurements obtained from each software
and dry mandibles were estimated.

STATA software (Stata Corp. 2003. Stata Statistical
Software: Release 8.0. College Station, TX: Stata Corpo-
ration, TX) was used.

Results

The ANOVA (Table 2) showed that the means of the
measurements performed in the four software systems
were not significantly different (P ¼ 0.958).

Comparing the radiographic measurements performed
in each of the four software systems to the dry mandibles,
all digital image measurements underestimated the defect
depth (t-test; P ø 0.000). The mean measurements were
5.07 (95% CI: 4.82–5.32), 5.06 (95% CI: 4.80–5.32), 5.01
(95% CI: 4.75–5.27), and 5.11 (95% CI: 4.84–5.37) and
6.67 (95% CI: 6.30–7.04) for Schick CDR DICOM
dedicated software, VixWin, Adobe Photoshop and
Image Tool non-dedicated software and dry mandibles,
respectively.

The bar graph (Figure 2) shows the difference between
the measurements from dry mandibles (gold standard) in
each one of the four imaging software systems.

Discussion

There were two important findings in this study: (1) that
measurements of simulated periodontal bone defects
obtained in one dedicated and three non-dedicated soft-
ware systems were not significantly different; (2) the bone
defect depth measurements in the digital radiographs were
underestimated compared with the measurements obtained
with the dry mandibles.

It is important to consider the nature and the facilities
provided in software systems developed for dental tasks
(Schick, VixWin and Image Tool) and for general
applications (Adobe Photoshop). For the latter, there is
no specific tool for image standardization, and the
standardization in the present study was made using a
known length of the aluminium step wedge. On the other
hand, Schick, VixWin and Image Tool software systems
provide a specific tool for standardization, which is
automatically prompted by the system in VixWin and
Image Tool when the image is opened on the monitor. In
the Schick system, it is necessary for the user to search for
this function. Thus, considering distance measurements,
the professional can choose each of these software systems,
both for dental tasks and for general applications.

For radiographic image interpretation using dedicated
and non-dedicated software, the results may be dependent
on the diagnostic tasks.

It seems that few studies have evaluated the interoper-
ability among software systems. One study assessed the
diagnostic accuracy of the morphology and position of
mandibular third molars in digital panoramic images using
the dedicated software and a general Picture Archiving and
Communication System (PACS) software. The results
were similar for the dedicated and the non-dedicated
software for all diagnostic tasks, except for the interpre-
tation of impacted third molars, for which the dedicated
software gave more accurate results.11 However, it is
difficult to compare our results with this study, since third
molar assessment is a different task from periodontal bone
defect measurements.

The results showed that digital radiographic measure-
ments using either of four software systems significantly
underestimated the defect depth measured in the dry

Table 1 General information about image software used in this study

Software name Company Website Dedicated Type Application

Schick DICOM Schick Technologies www.schicktech.com CDRw CMOS Dental
Vix Win PRO Gendex Dental System www.gendex-dental.com DenOptix/VisualiX PSP/CCD Dental
Image Tool 3.0 UTHSCSA ddsdx.uthscsa.edu/dig/itdesc.html – – Dental/biological
Photoshop 7.0 Adobe www.adobe.com – – General image processing

CCD, charge coupled device; CMOS-sensor, complementary metal oxide semiconductor; PSP, photo stimulable phosphor plate

Table 2 One-way analysis of variance for depth measurements data according to different software systems

Source of variation df Sum of squares Mean square F-value P-value

Programs 3 0.2039 0.0679 0.10 0.9580 ns
Residual 156 102.6754 0.6582
Total 159 102.8793

df, degrees of freedom; ns, non-significant

Bone defects in software systemsBone defects in software systems
G Scaf et al424

Dentomaxillofacial Radiology



mandible. The depth measured in radiographs was
approximately 25% lower than the dry mandible measure-
ments. Previous reports have demonstrated an under-
estimation ranging between 5.7% and 21% in bone
defects.12–14 The explanation for such underestimation

might be related to the two-wall periodontal defect that was
created, so that the remaining lingual wall was thick and
might have produced a blurred image. In a study conducted
by Pepelassi and Diamanti-Kipioti13 there was no reference
to the type of bone defect (one-, two- or three-wall
periodontal defects). In addition, we used a precision
digital calliper to measure the bone defects instead of the
periodontal probe used in previous reports.12,13

In conclusion, the periodontal bone defect measure-
ments using the Schick dedicated software CDR DICOM
for Windows v.3.5 and non-dedicated software Vix Win
2000 v.1.2, Image Tool 3.00 and Adobe Photoshop 7.0
were similar, while there was a considerable under-
estimation of bone defect depth measured in radiographs
compared to dry mandibles.

Acknowledgments

This study was supported by Grants from Fundação de Amparo à
Pesquisa do Estado de São Paulo #02/13326-7 and #02/13327-3.

Disclosure
The authors claim to have no financial interests in any company or
any of the products mentioned in this manuscript.

References

1. Hintze H, Wenzel A. Influence of the validation method on diagnostic
accuracy for caries. A comparison of six digital and two conventional
radiographic systems. Dentomaxillofac Radiol 2002; 31: 44–49.

2. Jacobsen JH, Hansen B, Wenzel A, Hintze H. Relationship between
histological and radiographic caries lesion depth measured in images
from four digital radiography systems. Caries Res 2004; 38: 34–38.

3. Furkart AJ, Dove SB, McDavid WD, Nummikoski P, Matteson S.
Direct digital radiography for the detection of periodontal bone
lesions. Oral Surg Oral Med Oral Pathol 1992; 74: 652–660.

4. Nair MK, Ludlow JB, Tyndall DA, Platin E, Denton G. Periodontitis
detection efficacy of film and digital images. Oral Surg Oral Med
Oral Pathol Oral Radiol Endod 1998; 85: 608–612.

5. Versteeg CH, Sanderink GC, van der Stelt PF. Efficacy of digital
intra-oral radiography in clinical dentistry. J Dent 1997; 25:
215–224.

6. Baker LC. Benefits of interoperability: a closer look at the estimates.
Health Aff (Millwood) 2005; 19: 22–25.

7. Benn DK, Bidgood WD Jr, Pettigrew JC Jr. An imaging standard for
dentistry. Extension of the radiology DICOM standard. Oral Surg
Oral Med Oral Pathol 1993; 76: 262–265.

8. Gotfredsen E, Wenzel A. A non-Dicom based system for distributing
and viewing digital images from multiple sources in dentistry. In:
Lemke HU, Vannier MW, Inamura K, Farman AG (eds). Computed
assisted radiology and surgery. Amsterdam: Excerpta Medica; 1999:
pp 964–967.

9. Wenzel A, Møystad A. Experience of Norwegian general dental
practitioners with solid state and storage phosphor detectors.
Dentomaxillofac Radiol 2001; 30: 203–208.

10. Berkhout WE, Sanderink GC, van der Stelt PF. A comparison of
digital and film radiography in Dutch dental practices assessed by
questionnaire. Dentomaxillofac Radiol 2002; 31: 93–99.

11. Benediktsdottir IS. Digital panoramic radiography for assessment of
mandibular third molars. [PhD thesis]. University of Aarhus, 2003.

12. Suomi JD, Plumbo J, Barbano JP. A comparative study of radiographs
and pocket measurements in periodontal disease evaluation.
J Periodontol 1968; 39: 311–315.

13. Pepelassi EA, Diamanti-Kipioti A. Selection of the most accurate
method of conventional radiography for the assessment of periodontal
osseous destruction. J Clin Periodontol 1997; 24: 557–567.

14. Akesson L, Hakansson J, Rohlin M. Comparison of panoramic and
intraoral radiography and pocket probing for the measurement of the
marginal bone level. J Clin Periodontol 1992; 19: 326–332.

Figure 2 Bar graph of the differences between the gold standard
measurements and radiographic measurements in four image softwares.
GS, gold standard; S, CDR DICOM v.3.5; VW, Vix Win 2000 v.1.2; A,
Adobe Photoshop 7.0; IT, Image Tool 3.0

Bone defects in software systemsBone defects in software systems
G Scaf et al 425

Dentomaxillofacial Radiology